The invention relates in general to spin transfer torque magnetic tunnel junctions (STT-MTJs), spin-transfer torque magnetic random access memories (STT-MRAMs) and, in particular, to perpendicularly magnetized STT-MTJs.
Perpendicularly magnetized devices, and in particular perpendicular STT-MRAMs, are known. Such devices are for instance discussed in the patents U.S. Pat. Nos. 7,602,000, 7,313,013, 7,943,399, 8,558,332, and U.S. Pat. No. 8,860,105.
It is generally believed that scaling challenges will prevent dynamic random-access memory (DRAM) and static random-access memory (SRAM) devices from functioning properly in coming technology nodes with the expected specifications. MRAM technology, in contrast, profits from two game-changing innovations.
The first innovation is based on the discovery that the magnetization of a nanoscale ferromagnet can be switched by a spin-polarized current [1,2]. An unpolarized current acquires a spin polarization by the electrons scattering in a ferromagnet. The spin imbalance that results can be transferred to the magnetic layer, eventually flipping its magnetization. Thus, writing to a memory cell—or switching its magnetization—no longer requires magnetic fields but can be achieved thanks to a short current pulse.
The second innovation is a materials engineering achievement. Originally, the magnetic layers in the magnetic tunnel junction (MTJ) were magnetized in-plane, driven by shape anisotropy. A few years ago, it was shown that magnetic materials like CoFeB can be tailored to exhibit “perpendicular” magnetization, i.e., where the magnetic orientation is out-of-plane, if the thickness of the element is small enough (typically 1 nm) [3,4,5]. A perpendicularly magnetized MTJ is advantageous over an in-plane magnetized MTJ because of its higher efficiency in switching the free layer, i.e. the minimal current needed for switching is reduced. The minimum current required for switching, often denoted by Ic0, depends on three material parameters: the damping constant α, the anisotropy barrier Eb, and the spin-transfer efficiency ζ. Low-power devices require Ic0 to be small, which, in turn, requires favorable material parameters. Efforts are thus usually invested in order to identify more favorable material parameters.
Attention is drawn to the following references:    [1] J. C. Slonczewski, J. Magn. Magn. Mater. 159, L1 (1996).    [2] L. Berger, Phys. Rev. B 54, 9353 (1996).    [3] M. Nakayama, T. Kai, N. Shimomura, M. Amano, E. Kitagawa, T. Nagase, M. Yoshikawa, T. Kishi, S. Ikegawa, and H. Yoda, J. Appl. Phys. 103, 07A710 (2008).    [4] S. Ikeda, K. Miura, H. Yamamoto, K. Mizunuma, H. D. Gan, M. Endo, S. Kanai, J. Hayakawa, F. Matsukura, and H. Ohno, Nature Mater. 9, 721 (2010).    [5] D. C. Worledge, G. Hu, D. W. Abraham, J. Z. Sun, P. L. Trouilloud, J. Nowak, S. Brown, M. C. Gaidis, E. J. O'Sullivan, and R. P. Robertazzi, Appl. Phys. Lett. 98, 022501 (2011).    [6] For the code, see A. Vanhaverbeke, OOMMF extension of spin-transfer torque terms for current-induced domain wall motion, 2008, http colon//www dot Zurich dot ibm dot com/st/magnetism/spintevolve dot html, see also, for an application, M. Najafi, B. Krüger, S. Bohlens, M. Franchin, H. Fangohr, A. Vanhaverbeke, R. Allenspach, M. Bolte, U. Merkt, D. Pfannkuche, D. P. F. Möller, and G. Meier, J. Appl. Phys. 105, 113914 (2009).    [7] C. Y. You, J. Magnetics 17, 73 (2012).